Epilepsy is a chronic brain disease that causes persistent and severe damage to the physical and mental health of patients. Daily effective prediction of epileptic seizures is crucial for epilepsy patients especially those with refractory epilepsy. At present, a large number of deep learning algorithms such as Convolutional Neural Networks and Recurrent Neural Networks have been used to predict epileptic seizures and have obtained better performance than traditional machine learning methods. However, these methods usually transform the Electroencephalogram (EEG) signal into a Euclidean grid structure. The conversion suffers from loss of adjacent spatial information, which results in deep learning models requiring more storage and computational consumption in the process of information fusion after information extraction. This study proposes a general Graph Convolutional Networks (GCN) model architecture for predicting seizures to solve the problem of oversized seizure prediction models based on exploring the graph structure of EEG signals. As a graph classification task, the network architecture includes graph convolution layers that extract node features with one-hop neighbors, pooling layers that summarize abstract node features; and fully connected layers that implement classification, resulting in superior prediction performance and smaller network size. The experiment shows that the model has an average sensitivity of 96.51%, an average AUC of 0.92, and a model size of 15.5 k on 18 patients in the CHB-MIT scalp EEG dataset. Compared with traditional deep learning methods, which require a large number of parameters and computational effort and are demanding in terms of storage space and energy consumption, this method is more suitable for implementation on compact, low-power wearable devices as a standard process for building a generic low-consumption graph network model on similar biomedical signals. Furthermore, the edge features of graphs can be used to make a preliminary determination of locations and types of discharge, making it more clinically interpretable.

The Xiaolangdi Dam is a key project for the control and development of the Yellow River. It bears the functions of flood control, controlling water and sediment in the lower reaches, ice prevention, industrial and agricultural water supply, power generation, and so on. Its safety is related to people’s life and property safety and local economic and social development. It is of great significance to carry out comprehensive and regular deformation monitoring for dams since the deformation is an important evaluation index for dam safety. Interferometric Synthetic Aperture Radar (InSAR) technology has been a rapidly evolving technology in the field of space geodesy in recent years. It offers advantages such as high monitoring precision, extensive coverage, and high monitoring point density, making it a powerful tool for monitoring deformations in hydraulic engineering projects. Based on Sentinel-1 data covering the Xiaolangdi Dam from September 2020 to November 2022, the PS-InSAR technique was used to obtain the surface deformation of the Xiaolangdi Dam, and reservoir water level data on image acquisition dates were obtained for joint analysis. The results show that there is a large deformation in the center of the dam crest of the Xiaolangdi Dam, while both sides of the slope and downstream dam foot are relatively stable. The time series deformation of the dam body is closely related to the reservoir water level change. When the water level increases, the dam body tends to deform downstream; when the water level decreases, the dam body tends to deform upstream. The deformation and water level of the Xiaolangdi Dam exhibit a clear negative correlation. There is no significant cumulative deformation on the dam slopes or at the base of the dam. However, cumulative deformation occurs over time in the central area of the dam’s crest. The deformation process at the central area of the dam’s crest follows a continuous and non-disruptive pattern, which is consistent with the typical deformation behavior of the Xiaolangdi earth–rock dam structure. Therefore, it is judged that the current deformation of the Xiaolangdi Dam does not impact the safe operation of the dam. InSAR technology enables the rapid acquisition of high-precision, high-density deformation information on the surfaces of reservoir dams. With an increasing number of radar satellites in various frequency bands, such as Sentinel-1 and TerraSAR-X, there is now an ample supply of available data sources for InSAR applications. Consequently, InSAR technology can be extended to routine monitoring applications for reservoir dam deformations, especially for small and medium-sized reservoirs that may not be equipped with ground measurement tools like GNSS. This holds significant importance and potential for enhancing the safety monitoring of such reservoirs.